Trans. zool. SOC.Lond. (1978) 34, 263-345 The phylogeny of the Charadriiformes (Aves): a new estimate using the method of character compatibility analysis JOSEPHG. STRAUCH,JR. Museum of Zoology, The University of Michigan, Ann Arbor, Michigan 48109, U.S.A.* (Accepted I2 October 1977) (With 36 figures in the text) Published for THE ZOOLOGICAL SOCIETY OF LONDON bY ACADEMIC PRESS * Present address: Science Applications Inc., 2760 29th St., Suite 209, Boulder, Colorado 80302, USA. COPYRIGHT 0 1978 BY THE ZOOLOGICAL SOCIETY OF LONDON REGENT'S PARK, LONDON, NWl4RY The phylogeny of the Charadriiformes (Aves): a new estimate using the method of character compatibility analysis Character state trees were devised for 70 mainly skeletal characters of 227 speciesof charadrii- form birds. Character compatibility analysis (described herein) was used to determine the largest sets of mutually compatible characters in the data set. Largest sets of mutually compatible characters were chosen as the best estimators of the phylogenetic history of the order, Smaller and smaller monophyletic groups (as identified by previous analysis of the next larger monophyletic group) were analysed to find locally largest sets of mutually compatible characters until the cladistic information in the data set was exhausted. The results of these analyses indicate that the Charadriiformes consist of three phyletic lines (here treated as suborders) : the Scolopaci, the Charadrii and the Alcae. The Scolopaci consist of the birds usually included in the families Jacanidae, Rostratulidae, Scolopacidae, Phalaropodidae and Thinocoridae. The Charadrii consist of two major phyletic branches: one leading to the Lari and the other to the line discussed below. The second branch of the Charadrii gives rise to five lineages: one which leads to Drornas, a second to Phviunellus and Chionis, a third to Pluviunus and theBurhinidae, a fourth to the Glareolidae, and a fifth to the plovers, lapwings, oystercatchers, Ibisbill, avocets, and stilts. Evidence which supports these findings, that which contradicts them, and relationships in need of further study are discussed. CONTENTS Page Introduction . .. .. .. .. .. .. .. .. .. .. 269 The method of character compatibility. ... .. .. .. .. *. .. 27 1 Characters.. .. .. .. .. .. .. .. .. .. .. .. 279 Skull characters . .. .. .. .. .. .. .. .. .. 289 Mandible characters. ... .. .. .. .. .. .. .. .. 301 Head and neck muscle characters .... .. .. .. .. *. .. 304 Axial skeleton characters . .. .. .. .. .. .. *. *. 304 Pectoral girdle characters . *. .. .. .. .. .. .. .. 306 Wing characters . .. .. .. .. .. .. .. .. .. 312 Synsacrum characters . .. .. .. .. .. .. *. .. 314 Hind limb characters .. .. .. .. .. .. .. .. .. 318 Results . .. .. .. .. .. .. .. .. .. .. .. 322 The Charadriiformes .. .. .. .. .. .. .. .. .. 322 The suborders of the Charadriiformes . .. .. .. .. .. *. 323 The Scolopaci . .. .. .. .. .. .. .. .. .. 323 The Rostratulidae, Scolopacidae, Phalaropodidae, and Thinocoridae *. 325 The Rostratulidae, Scolopacidae, and Phalaropodidae . .. .. 326 The Charadrii . .. .. .. .. .. .. .. .. .. 327 The plovers and their allies . .. .. .. .. .. .. 328 Analysis of the Charadrii using hypothetical ancestors . .. .. .. .. 328 Summary of the cladistic relationships among the Charadriiformes . .. *. 330 Discussion.. .. .. .. .. .. .. .. .. .. .. .. 333 Acknowledgements . .. .. .. .. .. .. .. .. .. 341 References , . .. .. .. .. .. .. .. .. .. .. .. 342 PHYLOGENY OF THE CHARADRIIFORMES 269 INTRODUCTION Much of the literature of systematics deals with the identification of characters which are good estimators of phylogenetic history. Early systematists had little more than their own insights to help them choose the characters which best indicate relationships. The stability of much of zoological classification is testimony to their good judgment in their choices. Their methods, however, have made it difficult or impossible for others to follow or repeat the steps from observations of specimens to the statements of relationship among taxa. The subjectivity of the intuitive method has led many systematists to reject classical phylogenetic studies and to accept phenetic methods (Sneath & Sokal, 1972). Pheneticists argue that classifications should be based on overall resemblance without regard to phylogeny. Those interested in estimating phylogenetic history, on the other hand, have long held that some characters are better than others for reconstructing phylogenies and that the major problem is to identify them. In recent years there has been a theoretical and methodological revolution in the estimation of phylogenetic relationships (Camin & Sokal, 1965; Hennig, 1966; Fitch & Margoliash, 1967; Kluge & Farris, 1969; Estabrook, 1972). Few of these new ideas, however, have been used in avian systematics (Selander, 1971 ; Cracraft, 1972). In this study I have used the method of character compatibility (Estabrook, 1972; McMorris, 1975; Estabrook, Strauch & Fiala, 1977) to estimate the branching patterns (cladistic relationships) of the phylogenetic history of the Charadriiformes. This method is based on current evolutionary theory, and its methods of analysis have been shown to be mathematically exact (Estabrook, Johnson & McMorris, 1975, 1976~7,b). Although the terminology and methodology of this method might at first seem foreign to many systematists, in truth it merely formulates in mathematical terms traditional systematic practices such as using the largest set of characters which agree to define taxonomic goups and treating some characters as more important indicators of relationships in some groups than in others. Unlike the Wagner Tree parsimony methods (Kluge & Farris, 1969) and the methods used by the followers of Hennig (e.g. Cracraft, 1974), character com- patibility analysis allows precise and objective identification of the best characters used in a study. The relationship between the method of character compatibility and traditional methods is discussed in more detail by Estabrook, Strauch & Fiala (1977). The avian order Charadriiformes includes birds commonly known as waders, gulls, terns and auks, as well as several less familiar types. This order is especially suitable for a phylogenetic study because it has given rise to many diverse types, members of which are found in almost all terrestrial and aquatic habitats, and because there have been several recent extensive studies of their systematics. All the birds currently included in the Charadriiformes were first grouped together by Huxley (1867) on the basis of skull characteristics. The major debates among bird systematists concerning charadriiform birds since Huxley’s work have centred around several problems : possible relationships among the Laridae, Procellariiformes, Gaviidae, 270 J. G. STRAUCH, JR and Alcidae (Sclater, 1880; Shufeldt, 1891), possible relationships among the Gruiformes and Charadriiformes, especially between the Otididae and Burhinidae (Garrod, 1873; Sclater, 1880; Sharpe, 1891 ; Lowe, 1931a; Boetticher, 1934; Stresemann, 1959) and between the Rallidae and the Jacanidae (Forbes, 1881 ; Lowe, 1925; Stresemann, 1959); the relationships of the Thinocoridae (Seebohm, 1895; Shufeldt, 1891 ; Mathews & Iredale, 1921; Lowe, 1922, 1923; Hanke & Niethammer, 1955; Sibley, Corbin & Ahlquist, 1968), and the question of whether the Pteroclidae are members of the Charadriiformes (Huxley, 1868; Garrod, 1874; Sclater, 1880; Gadow, 1893; Beddard, 1898; Maclean, 1967, 1969; Stegmann, 1968, 1969; Olson, 1970). Between 1914 and 1933 Percy R. Lowe published a series of papers on the anatomy and relationships of the Charadriiformes, especially waders. Lowe’s contributions include descriptions of the anatomy of several previously unstudied forms and discussions of the states of several characters in many different species. His decisions regarding relationships, however, are suspect because he relied heavily on only a few characters, such as colour pattern, supraorbital grooves, and the morphology of the quadrato-tympanic articulation. Kozlova (1961) and her collaborator Yudin (1965) proposed a phylogeny of the Charad- riiformes based on their “ecologico-morphology” method. Using information on behaviour, ecology, and the morphology of the head and wing, they began with the assumption that the genus Plzrvialis represents the archetype of the Charadriiformes from which more specialized charadriiform groups radiated. While they provide much valuable information on charadriiform morphology, the vagueness of their methods for inferring relationships and their use of much fragmentary and anecdotal evidence make many of their conclusions of doubtful value. In a survey of 24 orders of birds Kitto & Wilson (1966) found that charadriiform birds have a unique S-malate dehydrogenase, the mobility of which is 55 :(,that of most other birds. Their results indicate that the Jacanidae and Burhinidae are members of the Charadriiformes while the Pteroclidae are not. Jehl (1968), like Lowe (1915a), surveyed the plumage patterns of the downy young of waders. His conclusions, based on subjective estimates of similarity of overall patterns and the relationships implied by them, are in need of an objective reevaluation. Burton (1974) in an extensive survey of the feeding apparatus of waders discussed some of the systematic implications of his findings. Ahlquist (1974) used the IFPA (isoelectric focusing in polyacrylamide) patterns of egg-white proteins and two-dimensional electro- phoresis maps of